如何通过 Django 中的模型属性之一对对象数组进行分组？

Posted 2023-02-24

【中文标题】如何通过 Django 中的模型属性之一对对象数组进行分组？

【英文标题】：How to group an array of objects by one of the model's attribute in Django?

【发布时间】：2021-08-09 18:11:11

【问题描述】：

我想做类似于此示例 How to group an array of objects by key 的操作，但我在 Django 中找不到 API Get 的解决方案。

这是我所拥有的一个例子：

[
    
        "id": 1,
        "nutrition_measurement_id": "1",
        "nutrition_type_id": "2",
        "feeding_time": "dinner"
    ,
    
        "id": 2,
        "nutrition_measurement_id": "2",
        "nutrition_type_id": "1",
        "feeding_time": "dinner"
    ,
    
        "id": 3,
        "nutrition_measurement_id": "3",
        "nutrition_type_id": "2",
        "feeding_time": "breakfast"
    ,
    
        "id": 4,
        "nutrition_measurement_id": "2",
        "nutrition_type_id": "1",
        "feeding_time": "breakfast"
    
]

这是我想要实现的示例：

 "dinner": [
 
   "id": 3,
   "nutrition_type_id": 2,
   "nutrition_measurement_id": 1
 ,

   "id": 3,
   "nutrition_type_id": 1,
   "nutrition_measurement_id": 2
 
 ],

"breakfast": [
 
   "id": 3,
   "nutrition_type_id": 2,
   "nutrition_measurement_id": 3
 ,

    "id": 5,
    "nutrition_type_id": 1,
    "nutrition_measurement_id": 4
 
]

这是我拥有的模型：

class HorseNutrition(models.Model):
    horse = models.ForeignKey(Horse, models.DO_NOTHING)
    nutrition_type = models.ForeignKey('NutritionType', models.DO_NOTHING)
    nutrition_measurement = models.ForeignKey('NutritionMeasurement', models.DO_NOTHING)
    feeding_time = models.CharField(max_length=10, blank=True, null=True)
    class Meta:
        managed = False
        db_table = 'horse_nutrition'

class NutritionMeasurement(models.Model):
    name = models.CharField(unique=True, max_length=30, blank=True, null=True)
    class Meta:
        managed = False
        db_table = 'nutrition_measurement'

    def __str__(self):
        return self.name
    def __unicode__(self):
        return self.name


class NutritionType(models.Model):
    name = models.CharField(unique=True, max_length=30, blank=True, null=True)
    class Meta:
        managed = False
        db_table = 'nutrition_type'

    def __str__(self):
        return self.name
    def __unicode__(self):
        return self.name

这是我的序列化程序

class HorseNutritionSerializer(serializers.ModelSerializer):

    nutrition_measurement_id = serializers.ReadOnlyField(source = 'nutrition_measurement.name')
    nutrition_type_id = serializers.ReadOnlyField(source = 'nutrition_type.name')
    
    class Meta: 
        model = HorseNutrition
        fields = ['id', 'nutrition_measurement_id', 'nutrition_type_id', 'feeding_time']

这是我的看法：

class HorseNutritionView(APIView):
    
    def get (self, request, format = None):
        #id = int(request.GET.get(self.lookup_url_kwarg))
        id = 1
        if id != None:
            queryset = HorseNutrition.objects.filter(horse_id = id)
            serializer = HorseNutritionSerializer(queryset, many = True)             
            return Response(serializer.data)
        else:
            return Response(status = status.HTTP_400_BAD_REQUEST)

【参考方案1】：

您可以使用自定义ListSerializer（我们将在您的序列化程序的Meta 中设置list_serializer_class）并覆盖其to_representation 并使用itertools.groupby 对您的结果进行分组：

from itertools import groupby

class HorseNutritionListSerializer(serializers.ListSerializer):
    def to_representation(self, data):
        ret = super().to_representation(data)
        return key: list(group) for key, group in groupby(ret, key=lambda x: x['feeding_time'])
    
    @property
    def data(self):
        ret = serializers.BaseSerializer.data.fget(self)
        return serializers.ReturnDict(ret, serializer=self)


class HorseNutritionSerializer(serializers.ModelSerializer):
    nutrition_measurement_id = serializers.ReadOnlyField(source = 'nutrition_measurement.name')
    nutrition_type_id = serializers.ReadOnlyField(source = 'nutrition_type.name')
    
    class Meta: 
        model = HorseNutrition
        fields = ['id', 'nutrition_measurement_id', 'nutrition_type_id', 'feeding_time']
        list_serializer_class = HorseNutritionListSerializer

此外，要正确完成此分组，您的查询集必须是有序的，您可以在视图中执行此操作：

class HorseNutritionView(APIView):
    
    def get (self, request, format = None):
        #id = int(request.GET.get(self.lookup_url_kwarg))
        id = 1
        if id != None:
            queryset = HorseNutrition.objects.filter(horse_id = id).order_by('feeding_time')
            serializer = HorseNutritionSerializer(queryset, many = True)             
            return Response(serializer.data)
        else:
            return Response(status = status.HTTP_400_BAD_REQUEST)

【讨论】：

这很有意义，但我不知道为什么它不起作用:(

@olczig 啊，我的错，查询集（当 many=True 时）由父序列化程序 ListSerializer 在内部序列化，因此我们需要覆盖它的 to_representation，我们将不得不使用它自定义ListSerializer 并将其设置为list_serializer_class。检查我的编辑。

现在我只有喂食时间列表

@olczig 喂食时间列表？您应该得到一个 dictionary，其中包含馈送时间作为键和序列化 HorseNutrition 的列表作为值，因为这就是我们返回的内容。

@olczig 我从回答这个问题中学到了很多:)，问题是ListSerializer.data 返回ReturnList(ret, serializer=self) 所以字典变成了一个列表。我也编辑了解决这个问题的答案（也许在视图中使用itertools.groupby 会更容易，但你可以通过这种方式学习新东西；））